DynoHubs: What light bulbs/LED emitters?



On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:
> In article
> <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com>,
>  Dan O <[email protected]> wrote:
>
> > On Mar 15, 2:09 pm, [email protected] wrote:

>
> > > With one exception, I've never been in a place where a legally lit
> > > cyclist is not sufficiently conspicuous.  I've satisfied myself
> > > literally hundreds of times that I'm even more visible at night
> > > than I am in daytime.

>
> > WTF?!?

>
> WTF do you mean "WTF?!?"  


WTF do you mean WTF be means WTF?

>A bicyclist at night with a headlight and
> taillight is more conspicuous than a cyclist riding during the day.  Way
> more contrast between a bright light and a dark background.


In total darkness, a condition that does not exist for most cycling
situations. This is a juvenile form of argument, Tim.

And even your basic argument is fatally undermined by the deficiency
that the light must be pretty strong to be seen at the same maximum
distance as a cyclist remains visiblein daylight

Andre Jute
Adult cycling club
 
On Mar 16, 10:07 am, Andre Jute <[email protected]> wrote:
> On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:
>
> > In article
> > <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com>,
> >  Dan O <[email protected]> wrote:

>
> > > On Mar 15, 2:09 pm, [email protected] wrote:

>
> > > > With one exception, I've never been in a place where a legally lit
> > > > cyclist is not sufficiently conspicuous.  I've satisfied myself
> > > > literally hundreds of times that I'm even more visible at night
> > > > than I am in daytime.

>
> > > WTF?!?

>
> > WTF do you mean "WTF?!?"  

>
> WTF do you mean WTF be means WTF?
>
> >A bicyclist at night with a headlight and
> > taillight is more conspicuous than a cyclist riding during the day.  Way
> > more contrast between a bright light and a dark background.

>
> In total darkness, a condition that does not exist for most cycling
> situations. This is a juvenile form of argument, Tim.
>
> And even your basic argument is fatally undermined by the deficiency
> that the light must be pretty strong to be seen at the same maximum
> distance as a cyclist remains visiblein daylight
>
> Andre Jute
> Adult cycling club


It really does depend on the light system. I was driving home the
other night, and a cyclist approached from the other direction with
two, forward facing flashing LED arrays -- one on the bars and one on
the helmet. These things were blinding. I thought I had stumbled in
to an alien landing. This sort of retina burning, visible-from-
outerspace display would be far more noticeable to a car enterning
traffic from a side street (the usual culprits in a night time
colision, IMO) than a rider in daylight. This system, however, far
exceeded any legal requirement -- for any conveyance, jet, boat, train
or space ship. --- Jay Beattie.
 
On Mar 14, 9:05 pm, Andre Jute <[email protected]> wrote:

> 3. Forget a dynopowered rear light of any kind. Even the best are
> dangerous to your health. The expensive BUMM ones are not watertight
> and the best of the rest, made by Basta (I have one they custom make
> for Gazelle but it is basically just an aesthetic variation of their
> best rear lamp) and by Spanninga (their Ultra; I have that one as
> well) are merely better waterproofed, not more illuminative. I keep
> them on my bikes simply because they came with the bikes. Even the
> best of the dynodriven rear lights are little glimmers that you can
> barely see across the street. None of the dyno-driven rear lights
> flash, because it is streng verboten to have flashing lights in
> Germany and The Netherlands, their prime markets.


Good advice. I had some long discussions with Dutch manufacturers at
the show about their products (and their lack of exporting to the
U.S.), and especially about lighting. I was very suprised to see some
new higher end Dutch commuter bikes with no dynamos at all, hub or
rim. They told me that there is a trend even in the Netherlands toward
battery powered lights at the mid-range for two reasons. First, the
rim dynamos are too unreliable in terms of wiring and in terms of
being damaged when the bicycles are parked, but the hub dynamos are
too expensive except at the very high end. Second, the lights are only
useful as "being seen" lamps at the relatively slow speeds on the
cycle paths.

> 4. Get a battery rear light. If you're rich, get a Dinotte rear light
> (ask Jay; he has one), if not a Cateye TL-LD1100, which is pricey
> enough. There is only one other taillight that is good enough for your
> life and that's the Trek Disco Inferno, which is no longer made. The
> Dinotte and the Cateye 1100 are *bright*, they cast very substantial
> light to the sides as well as the rear, and they flash. Those are the
> minimum requirements for good taillights, and they are the only ones
> who truly meet them. The Cateye 1100 is bright enough to be seen in
> bright sunlight; I use it as a daylight running lamp. It is supposed
> to last 200 hours on a set of 2 AA batteries; I don't know how long
> the batteries last in hours because I use rechargeables and swap them
> out every three or four months or so.
>


Other than folding bicycles, there were probably more new attempts at
LED lights than any other product at the show, including several LED
brake lights and turn signal lights, including some with wireless
transmitters from a switch on the bars But in all the show, the best
rear light remained the CatEye 1100. DiNotte was not at the show.

There actually is a decent tail light from Blackburn (in terms of
brightness and angle of view) but it suffers from using AAA batteries.

> 7. Or you might want to considering overvolting a single halogen lamp:
> you get far more light and you won't blow a Philips MR16 or MR11
> longlife unit --anyway, what do you care if you reduce a 3000 mtbf
> lamp to 1500 hours of life if you get nearly twice as much light? The
> trick is that you must be able to get them in the 6V versions to work
> with your dynohub, and the 6V MR16 or MR11 are not easy to find, at
> least not where I live.


Yeah, the poor man's HID system! Have you tried a 5W 6V MR16 on a 3W
dynohub?

"http://www.bulbtown.com/5W_6V_MR16_WITH_LENSE_GX5_3_BASE_p/43243.htm"

> 8. Or, in LEDs, you can fit as many low consumption LEDs as you can
> power. Each LED drops y volts, so the total must add up to what your
> dynohub produces or must be regulated. You might want to look into
> buckpucks to get the voltage right. Frankly, I wouldn't mess with LEDs
> unless I could get the latest and the best, together with some means
> of focusing the light correctly, and were also willing to sacrifice an
> existing set of lights with hefty, preferably cast ali, shells for
> cooling the LEDs. I looked into LEDs and decided that BUMM's Fly IQ
> (at the expensive end of their range, which is generally overpriced)
> would probably in the end cost less than messing around trying to make
> my own.


I was surprised to see so few LED based dynamo lights at the show,
because I incorrectly assumed that you could drive something like a
3W Cree LED from a dynamo hub. You can't. Most of the dynamo LED based
lights were three lower wattage LEDs, and not focused or collimated
all that well. Plus no manufacturer is apparantly willing to spend big
bucks on the bins of the Cree LED that are the most efficient.

The best LED headlamp appears to still be the SolidLight's 1203D.
Interestingly, tthey don't mention the LED type anywhere, nor the
wattage.

> 9. If you're cheap or poor, consider this. Plenty of RBT dickswingers
> will now weight in with how fabulous their BUMM Fly IQ is; I have one
> too and it is a good light. However. A couple of halogen 2.4W lamps --
> because that is what I had at the time of the test; 2x 3W lights would
> do better still -- made as much light as the Fly at any speed over a
> crawl and could be better arranged because the two lamps had different
> spreads.


There is a mistaken belief by many that an LED based light is
necessarily more efficient than a filament based lamp. This is untrue
for higher power lights. The measures taken to dissipate the heat from
a high power LED lamp are quite incredible. Plus an LED llamp is much
harder to lens properly. There are advantages, such as the longer life
of the LED compared to the filament based bulb, especially in a harsh
environment. You keep hearing how LEDs will soon catch up with HID in
terms of efficiency, and this may happen but it's not going to be
cheap. The LED manufacturers already charge a big premium for the
their most efficient product bins, and you see Cree based lights
specifying which bin the LEDs come from.

> 10. Lights are the last bicycle frontier. We hear a lot of talk from
> the technofreakies about how dynamo lights are now so much better than
> they were. But better isn't automatically good enough. The best dyno
> front light is still only nearly as good as a 10W MR11 battery light
> -- whereas I don't feel comfortable on any aspect of lighting (being
> seen, having my space respected, seeing) with anything less than about
> 25W divided between two lamps. YMMV, of course.


Personally I find MR11 based lamps a waste, because the larger
reflector of the MR16 is much more efficient. In fact the sealed beam
12 volt lamps are becoming one of my favorite halogen lamps because of
the large reflector, light weight, and the lack of a need to build any
funky enclosure.

> 11. In summary: I recommend the Cateye TL-LD1100 battery rear light,
> and two cheap BUMM halogen lights driven off the dynamo at the front
> with a homemade switch, supplemented in case of regular commuting or
> any strenuous riding circumstances by a rechargeable battery front
> light set .


Good advice. I'm finding that for many short, slow rides to the store
that the dynamo lights are sufficient. It's for commuting, especially
at relatively high speed, that the higher power lights are necessary
for optimal safety.

> 13. Good strong lights are useful in daylight too. The flashing Cateye
> 1100 persuades a lot of people to slow behind me and to give me a
> wider berth than they did before I fitted that light.


Try a Flash Flag. See "http://www.flashback.ca/bicycle.html". You can
probably make something similar. I use these on most of our fleet of
bikes, but I'm working on something better. I don't like the lack of a
good breakaway mechanism. Some sort of replaceable, cheap mechanism is
needed. I was hoping to see something similar to the Flash Flag at the
bike show, but there was nothing.
 
Per SMS:
>Flash Flag. See "http://www.flashback.ca/bicycle.html".


Somebody marketed something like that back in the late
seventies/early eighties called "The Sting".

AFIK, it didn't go anywhere.
--
PeteCresswell
 
On Mar 16, 7:55 pm, Jay Beattie <[email protected]> wrote:
> On Mar 16, 10:07 am, Andre Jute <[email protected]> wrote:
>
>
>
> > On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:

>
> > > In article
> > > <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com>,
> > >  Dan O <[email protected]> wrote:

>
> > > > On Mar 15, 2:09 pm, [email protected] wrote:

>
> > > > > With one exception, I've never been in a place where a legally lit
> > > > > cyclist is not sufficiently conspicuous.  I've satisfied myself
> > > > > literally hundreds of times that I'm even more visible at night
> > > > > than I am in daytime.

>
> > > > WTF?!?

>
> > > WTF do you mean "WTF?!?"  

>
> > WTF do you mean WTF be means WTF?

>
> > >A bicyclist at night with a headlight and
> > > taillight is more conspicuous than a cyclist riding during the day.  Way
> > > more contrast between a bright light and a dark background.

>
> > In total darkness, a condition that does not exist for most cycling
> > situations. This is a juvenile form of argument, Tim.

>
> > And even your basic argument is fatally undermined by the deficiency
> > that the light must be pretty strong to be seen at the same maximum
> > distance as a cyclist remains visiblein daylight

>
> > Andre Jute
> > Adult cycling club

>
> It really does depend on the light system.  I was driving home the
> other night, and a cyclist approached from the other direction with
> two, forward facing flashing LED arrays -- one on the bars and one on
> the helmet.  These things were blinding.  I thought I had stumbled in
> to an alien landing. This sort of retina burning, visible-from-
> outerspace display would be far more noticeable to a car enterning
> traffic from a side street (the usual culprits in a night time
> colision, IMO) than a rider in daylight.  This system, however, far
> exceeded any legal requirement -- for any conveyance, jet, boat, train
> or space ship. --- Jay Beattie.


I couldn't agree with you more, Jay. A blinding light is as
counterproductive as an inadequate light. Bicycle lights, front and
rear, should be adequate to the task but not a hazard to other road
users. But on RBT it must often seem to you that you are the last
moderate here; I keep stumbling into entrenched prejudice that leads
to vicious responses on helmets and lights, not to mention motor
pacing... -- Andre Jute
 
On Mar 16, 2:52 pm, "(PeteCresswell)" <[email protected]> wrote:
> Per SMS:
>
> >Flash Flag. See "http://www.flashback.ca/bicycle.html".

>
> Somebody marketed something like that back in the late
> seventies/early eighties called "The Sting".
>
> AFIK, it didn't go anywhere.
> --
> PeteCresswell


Too bad. I find the Flash Flag to be very effective in having
motorists give me more room. Whether it's because they see me better,
or because they just don't want to risk scratching their vehicle, it's
a simple device that achieves its purpose, including in the daytime
where only a few LED flashers are bright enough to be effective.
 
On Mar 14, 8:05 pm, bob prohaska's usenet account <[email protected]>
wrote:
> "(PeteCresswell)" <[email protected]> wrote:
> > Per Hank:
> >>I was a liberal arts major, so I bought a Lumotec IQ Fly , crimped the
> >>spade connectors, and plugged it into my SON. Piece of cake.

>
> > Having said all that.... Thanks for the terminology. Googling
> > "Lumotec" led right to Peter White's page dedicated to generator
> > lighting.
> >http://www.peterwhitecycles.com/schmidt-headlights.asp

>
> Bike generators match very well to light emitting
> diodes.


The problem wiht LEDs is that the AC, non-regulated voltage out of the
dynamo isn't well suited to driving LEDs without some extra
electronics to convert the AC to DC and to keep the current and
voltage constant. To achieve this at low cost and at high efficiency
is not yet possible. This is why the few good dynamo LED lights are so
expensive.

A filament bulb requires only some over-voltage protection, doesn't
care about AC or DC, and lends itself to more efficient optics.

I wish that there would be some higher power dynamos, even at the
expense of more drag and more weight. All that's needed is enough
power to keep a 3W Cree LED, battery powered light, charged. A 4 cell
Cree LED light that automatically switched between two sets of two
cells (operating on one while charging the other from a dynamo) would
be one solution that would be possible with a slightly higher power
dynamo, and would give the best of both worlds. But you're talking
about a $100 headlight by the time it's at the retail level in
volume.
 
In article
<[email protected]>,
Jay Beattie <[email protected]> wrote:

> On Mar 16, 10:07 am, Andre Jute <[email protected]> wrote:
> > On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:
> >
> > > In article
> > > <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com
> > > >,  Dan O <[email protected]> wrote:

> >
> > > > On Mar 15, 2:09 pm, [email protected] wrote:

> >
> > > > > With one exception, I've never been in a place where a
> > > > > legally lit cyclist is not sufficiently conspicuous.  I've
> > > > > satisfied myself literally hundreds of times that I'm even
> > > > > more visible at night than I am in daytime.

> >
> > > > WTF?!?

> >
> > > WTF do you mean "WTF?!?"  

> >
> > WTF do you mean WTF be means WTF?
> >
> > >A bicyclist at night with a headlight and
> > > taillight is more conspicuous than a cyclist riding during the
> > > day.  Way more contrast between a bright light and a dark
> > > background.

> >
> > In total darkness, a condition that does not exist for most cycling
> > situations. This is a juvenile form of argument, Tim.
> >
> > And even your basic argument is fatally undermined by the
> > deficiency that the light must be pretty strong to be seen at the
> > same maximum distance as a cyclist remains visiblein daylight


Try testing this in reality rather than in your imagination. The
reverse of your position is true.

> It really does depend on the light system. I was driving home the
> other night, and a cyclist approached from the other direction with
> two, forward facing flashing LED arrays -- one on the bars and one on
> the helmet. These things were blinding. I thought I had stumbled in
> to an alien landing. This sort of retina burning, visible-from-
> outerspace display would be far more noticeable to a car enterning
> traffic from a side street (the usual culprits in a night time
> colision, IMO) than a rider in daylight. This system, however, far
> exceeded any legal requirement -- for any conveyance, jet, boat,
> train or space ship. --- Jay Beattie.


Which some participants in these discussions would still find to be
deficient lighting in some way.
 
On Mar 16, 9:00 pm, SMS <[email protected]> wrote:
> On Mar 14, 9:05 pm, Andre Jute <[email protected]> wrote:
>
> > 3. Forget a dynopowered rear light of any kind. Even the best are
> > dangerous to your health. The expensive BUMM ones are not watertight
> > and the best of the rest, made by Basta (I have one they custom make
> > for Gazelle but it is basically just an aesthetic variation of their
> > best rear lamp)  and by Spanninga (their Ultra; I have that one as
> > well) are merely better waterproofed, not more illuminative. I keep
> > them on my bikes simply because they came with the bikes. Even the
> > best of the dynodriven rear lights are little glimmers that you can
> > barely see across the street. None of the dyno-driven rear lights
> > flash, because it is streng verboten to have flashing lights in
> > Germany and The Netherlands, their prime markets.

>
> Good advice. I had some long discussions with Dutch manufacturers at
> the show about their products (and their lack of exporting to the
> U.S.), and especially about lighting. I was very suprised to see some
> new higher end Dutch commuter bikes with no dynamos at all, hub or
> rim. They told me that there is a trend even in the Netherlands toward
> battery powered lights at the mid-range for two reasons. First,  the
> rim dynamos are too unreliable in terms of wiring and in terms of
> being damaged when the bicycles are parked, but the hub dynamos are
> too expensive except at the very high end. Second, the lights are only
> useful as  "being seen" lamps at the relatively slow speeds on the
> cycle paths.


That's a retrograde step. Both my Dutch bicycles have dynamo-driven
front lights and battery rear lights, a very common arrangement at the
high end, the reason being the problem (aesthetics, reliability) of
wiring a dynamo-driven light at the back. But conceptually, once you
accept that the dynamo lights are being-seen lights, or seeing lights
only in undemanding situations, and that for any demanding use
(including by people with a true value of their own lives) such lights
must be supplemented by battery lights, it makes sense to have dyno-
drive lights front *and* back as backups in case the batteries run
out. I'm sure that if they put their minds to it the bike makers or
lamp makers could solve the problem right pronto. (However, I notice
that Shimano is no longer supplying either the front lamp they used to
list, nor the switch they used to list for dynohubs; I wonder if the
glow has gone off the dynohub market for Shimano.)

> > 4. Get a battery rear light. If you're rich, get a Dinotte rear light
> > (ask Jay; he has one), if not a Cateye TL-LD1100, which is pricey
> > enough. There is only one other taillight that is good enough for your
> > life and that's the Trek Disco Inferno, which is no longer made. The
> > Dinotte and the Cateye 1100 are *bright*, they cast very substantial
> > light to the sides as well as the rear, and they flash. Those are the
> > minimum requirements for good taillights, and they are the only ones
> > who truly meet them. The Cateye 1100 is bright enough to be seen in
> > bright sunlight; I use it as a daylight running lamp. It is supposed
> > to last 200 hours on a set of 2 AA batteries; I don't know how long
> > the batteries last in hours because I use rechargeables and swap them
> > out every three or four months or so.

>
> Other than folding bicycles, there were probably more new attempts at
> LED lights than any other product at the show, including several LED
> brake lights and turn signal lights, including some with wireless
> transmitters from a switch on the bars But in all the show, the best
> rear light remained the CatEye 1100.  DiNotte was not at the show.
>
> There actually is a decent tail light from Blackburn (in terms of
> brightness and angle of view) but it suffers from using AAA batteries.


That makes all of *three* generally available good tail lights... A
shameful situation, I think.

There's also a rather ugly front light from Blackburn that I might us
as a front flasher. But first I want to look into making something
like the BUMM Fly IQ flash, so it can run off the dynohub and be there
in cast I want to use it in steady mode. I'm also wondering if I can
make the Cateye 1100 work off the dynohub for the rear.

> > 7. Or you might want to considering overvolting a single halogen lamp:
> > you get far more light and you won't blow a Philips MR16 or MR11
> > longlife unit --anyway, what do you care if you reduce a 3000 mtbf
> > lamp to 1500 hours of life if you get nearly twice as much light? The
> > trick is that you must be able to get them in the 6V versions to work
> > with your dynohub, and the 6V MR16 or MR11 are not easy to find, at
> > least not where I live.

>
> Yeah, the poor man's HID system! Have you tried a 5W 6V MR16 on a 3W
> dynohub?
>
> "http://www.bulbtown.com/5W_6V_MR16_WITH_LENSE_GX5_3_BASE_p/43243.htm"


Thanks for the link. They don't deliver to Europe though. The reason I
haven't tried 6V MR16 is that I can't find any locally or deliverable
for a reasonable price, and in fact have several perfectly good (of
their kind, for their purpose) 6V dyno front lights made by Basta,
Spanninga and BUMM.

> > 8. Or, in LEDs, you can fit as many low consumption LEDs as you can
> > power. Each LED drops y volts, so the total must add up to what your
> > dynohub produces or must be regulated. You might want to look into
> > buckpucks to get the voltage right. Frankly, I wouldn't mess with LEDs
> > unless I could get the latest and the best, together with some means
> > of focusing the light correctly, and were also willing to sacrifice an
> > existing set of lights with hefty, preferably cast ali, shells for
> > cooling the LEDs. I looked into LEDs and decided that BUMM's Fly IQ
> > (at the expensive end of their range, which is generally overpriced)
> > would probably in the end cost less than messing around trying to make
> > my own.

>
> I was surprised to see so few LED based dynamo lights at the show,
> because  I incorrectly assumed that you could drive something like a
> 3W Cree LED from a dynamo hub. You can't.


No free lunch? I assume it consumes too much current.

> Most of the dynamo LED based
> lights were three lower wattage LEDs, and not focused or collimated
> all that well. Plus no manufacturer is apparantly willing to spend big
> bucks on the bins of the Cree LED that are the most efficient.
>
> The best LED headlamp appears to still be the SolidLight's 1203D.


That's *very* expensive, Steven. I looked into the SolidLight and was
tempted, but it will be outmoded before I've even run it in, and then
my money will be wasted.

> Interestingly, tthey don't mention the LED type anywhere, nor the


Luxeon 3W from a select bin, I imagine.

> > 9. If you're cheap or poor, consider this. Plenty of RBT dickswingers
> > will now weight in with how fabulous their BUMM Fly IQ is; I have one
> > too and it is a good light. However. A couple of halogen 2.4W lamps --
> > because that is what I had at the time of the test; 2x 3W lights would
> > do better still -- made as much light as the Fly at any speed over a
> > crawl and could be better arranged because the two lamps had different
> > spreads.

>
> There is a mistaken belief by many that an LED based light is
> necessarily more efficient than a filament based lamp. This is untrue
> for higher power lights. The measures taken to dissipate the heat from
> a high power LED lamp are quite incredible. Plus an LED llamp is much
> harder to lens properly. There are advantages, such as the longer life
> of the LED compared to the filament based bulb, especially in a harsh
> environment. You keep hearing how LEDs will soon catch up with HID in
> terms of efficiency, and this may happen but it's not going to be
> cheap. The LED manufacturers already charge a big premium for the
> their most efficient product bins, and you see Cree based lights
> specifying which bin the LEDs come from.
>
> > 10. Lights are the last bicycle frontier. We hear a lot of talk from
> > the technofreakies about how dynamo lights are now so much better than
> > they were. But better isn't automatically good enough. The best dyno
> > front light is still only nearly as good as a 10W MR11 battery light
> > -- whereas I don't feel comfortable on any aspect of lighting (being
> > seen, having my space respected, seeing) with anything less than about
> > 25W divided between two lamps. YMMV, of course.

>
> Personally I find MR11 based lamps a waste, because the larger
> reflector of the MR16 is much more efficient. In fact the sealed beam
> 12 volt lamps are becoming one of my favorite halogen lamps because of
> the large reflector, light weight, and the lack of a need to build any
> funky enclosure.


I looked into those garden lighting decorator types of sealed beams
you recommend but again it was a supply problem.

> > 11. In summary: I recommend the Cateye TL-LD1100 battery rear light,
> > and two cheap BUMM halogen lights driven off the dynamo at the front
> > with a homemade switch, supplemented in case of regular commuting or
> > any strenuous riding circumstances by a rechargeable battery front
> > light set .

>
> Good advice. I'm finding that for many short, slow rides to the store
> that the dynamo lights are sufficient. It's for commuting, especially
> at relatively high speed, that the higher power lights are necessary
> for optimal safety.


I really hoped that a hub dynamo would be so much better than sidewall
dynamos but it is not so: the light output in my kinds of mild use is
no higher. The truth is that a hub dynamo with even the best lamps
designed for it doesn't make adequate light for any of my nighttime
rides -- I might turn from the shop across the unlit graveyard, or
from a garage which sometimes has a veteran car into the rough tracks
on the estate of a friend, and in both places the spread of light
from dynohub lights, and the difficulty of starting up again after you
stop, make the ride into an unnecessarily stressful adventure.

> > 13. Good strong lights are useful in daylight too. The flashing Cateye
> > 1100 persuades a lot of people to slow behind me and to give me a
> > wider berth than they did before I fitted that light.

>
> Try a Flash Flag. See "http://www.flashback.ca/bicycle.html". You can
> probably make something similar. I use these on most of our fleet of
> bikes, but I'm working on something better. I don't like the lack of a
> good breakaway mechanism. Some sort of replaceable, cheap mechanism is
> needed. I was hoping to see something similar to the Flash Flag at the
> bike show, but there was nothing.


I think the Cateye 1100 at last does the business, even in daylight;
if Cateye's next model is as much of an improvement over the 1100 as
the 1100 was over the 1000, I will buy that one too. A guy with a 1000
was passing in the winter here at dusk, very poor visibility. He
immediately noticed my 1100 was stronger. We fitted fresh batteries
from my bulk pack, parked the bikes on the hard shoulder with the
lights flashing and walked back. The 1100 was visible for more than an
additional 150 paces, at which point a curve in the road intervened.
This fellow, a Brit, had thought the 1000 a revelation, very good
indeed on cycling hols he takes all over the world every few weeks
when he can get a cheap flight, but he said he would buy the 1100 next
time he was in a shop. He was the one that pointed out to me that the
key is that the bike lights should still be noticeable under street
lights and in traffic; that a test in pitch dark is an engineering
test but useless as a utility test. My Spanninga Ultra, highly thought
of in The Netherlands, recommended in a Fietserbond test, was on the
same bike as the Cateye 1100 and didn't pass our impromptu test.

Andre Jute
I wouldn't want to be on even a cager's conscience
 
>>> Per Hank:
>>>> I was a liberal arts major, so I bought a Lumotec IQ Fly , crimped the
>>>> spade connectors, and plugged it into my SON. Piece of cake.


>> "(PeteCresswell)" <[email protected]> wrote:
>>> Having said all that.... Thanks for the terminology. Googling
>>> "Lumotec" led right to Peter White's page dedicated to generator
>>> lighting.
>>> http://www.peterwhitecycles.com/schmidt-headlights.asp


> bob prohaska's usenet account <[email protected]> wrote:
>> Bike generators match very well to light emitting
>> diodes.


SMS wrote:
> The problem wiht LEDs is that the AC, non-regulated voltage out of the
> dynamo isn't well suited to driving LEDs without some extra
> electronics to convert the AC to DC and to keep the current and
> voltage constant. To achieve this at low cost and at high efficiency
> is not yet possible. This is why the few good dynamo LED lights are so
> expensive.
>
> A filament bulb requires only some over-voltage protection, doesn't
> care about AC or DC, and lends itself to more efficient optics.
>
> I wish that there would be some higher power dynamos, even at the
> expense of more drag and more weight. All that's needed is enough
> power to keep a 3W Cree LED, battery powered light, charged. A 4 cell
> Cree LED light that automatically switched between two sets of two
> cells (operating on one while charging the other from a dynamo) would
> be one solution that would be possible with a slightly higher power
> dynamo, and would give the best of both worlds. But you're talking
> about a $100 headlight by the time it's at the retail level in
> volume.


I thought I might throw out my bikes and just sit down and cry because
the world is imperfect. Instead, I just ride with normal dynamo lights.
Installed once and never given a moment's thought since.
--
Andrew Muzi
www.yellowjersey.org
Open every day since 1 April, 1971
 
On Mar 16, 4:24 pm, Andre Jute <[email protected]> wrote:

> That's a retrograde step. Both my Dutch bicycles have dynamo-driven
> front lights and battery rear lights, a very common arrangement at the
> high end, the reason being the problem (aesthetics, reliability) of
> wiring a dynamo-driven light at the back. But conceptually, once you
> accept that the dynamo lights are being-seen lights, or seeing lights
> only in undemanding situations, and that for any demanding use
> (including by people with a true value of their own lives) such lights
> must be supplemented by battery lights, it makes sense to have dyno-
> drive lights front *and* back as backups in case the batteries run
> out. I'm sure that if they put their minds to it the bike makers or
> lamp makers could solve the problem right pronto. (However, I notice
> that Shimano is no longer supplying either the front lamp they used to
> list, nor the switch they used to list for dynohubs; I wonder if the
> glow has gone off the dynohub market for Shimano.)


I was talking to someone from Sturmey Archer at Taipei Cycle a few
days ago, and I was looking at their combination drum brake/dynamo. I
asked if they were going to do a dynamo-only, having no interest in
drum brakes, but he said that there was just not enough demand, and
not enough money in it. While the dynohubs sold after-market fetch
high prices, the bicycle manufacturers pay probably 1/10th of the
retail price for a dynamo, and even in Europe the dynamos are losing
favor. Ironically it's in the U.S. where there is some renewed
interest in dynamos from the likes of the Breezer bikes, and a few
other commuter bikes. Dahon offers the Biologic Joule hub, apparently
made only for them.

I think that one of the reasons that the dynamo lights are losing
favor is because LEDs aren't well suited for making cheap headlights
for dynamos, but are very well suited for cheap battery lights. Still,
the good LED bicycle lights, using the best binned LEDs, are not cheap
enough for the mass market. There's a limited demand for DiNotte
lights at those prices. DiNotte wants to buy one of my domain names,
but I need to wait until they sell enough lights to meet my price!

I was very impressed with one company's Cree LED bicycle light and
flashlight. They did very good optics for the bicycle light, but in
fact they tried too hard! One of the big advantages of a standard Cree
LED flashlight is that it _does illuminate off to the sides, and not
just directly in front of the bicycle. Contrary to what some people
believe, it's not illumuniting the sky!

The highlight of my trip was meeting Joe Breeze as he rode in from his
hotel on a Dahon, lent to him by Josh DaHon. Of course I made an idiot
out of myself before I realized who he was! The mindset was the same
anyway, stay out by the main train station and MRT in a two star
hotel, where it's much more interesting, as well as much less
expensive than the 5 star hotels where most of the bicycle execs stay.
He asked if my hotel had a window, and said that his window looked out
on nothing but the window of the room across.

Well gotta go pack for the trip back.
 
On Mar 16, 4:15 pm, Tim McNamara <[email protected]> wrote:
> In article
> <[email protected]>,
>  Jay Beattie <[email protected]> wrote:
>
>
>
>
>
> > On Mar 16, 10:07 am, Andre Jute <[email protected]> wrote:
> > > On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:

>
> > > > In article
> > > > <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com
> > > > >,  Dan O <[email protected]> wrote:

>
> > > > > On Mar 15, 2:09 pm, [email protected] wrote:

>
> > > > > > With one exception, I've never been in a place where a
> > > > > > legally lit cyclist is not sufficiently conspicuous.  I've
> > > > > > satisfied myself literally hundreds of times that I'm even
> > > > > > more visible at night than I am in daytime.

>
> > > > > WTF?!?

>
> > > > WTF do you mean "WTF?!?"  

>
> > > WTF do you mean WTF be means WTF?

>
> > > >A bicyclist at night with a headlight and
> > > > taillight is more conspicuous than a cyclist riding during the
> > > > day.  Way more contrast between a bright light and a dark
> > > > background.

>
> > > In total darkness, a condition that does not exist for most cycling
> > > situations. This is a juvenile form of argument, Tim.

>
> > > And even your basic argument is fatally undermined by the
> > > deficiency that the light must be pretty strong to be seen at the
> > > same maximum distance as a cyclist remains visiblein daylight

>
> Try testing this in reality rather than in your imagination.  The
> reverse of your position is true.
>
> > It really does depend on the light system.  I was driving home the
> > other night, and a cyclist approached from the other direction with
> > two, forward facing flashing LED arrays -- one on the bars and one on
> > the helmet.  These things were blinding.  I thought I had stumbled in
> > to an alien landing. This sort of retina burning, visible-from-
> > outerspace display would be far more noticeable to a car enterning
> > traffic from a side street (the usual culprits in a night time
> > colision, IMO) than a rider in daylight.  This system, however, far
> > exceeded any legal requirement -- for any conveyance, jet, boat,
> > train or space ship. --- Jay Beattie.

>
> Which some participants in these discussions would still find to be
> deficient lighting in some way.- Hide quoted text -


I would find it deficient lighting if I were riding the bike. I don't
like riding with front flashing lights. I mean really, it's a road
not a disco.-- Jay Beattie.
 
On Mar 16, 4:24 pm, Andre Jute <[email protected]> wrote:

> That makes all of *three* generally available good tail lights... A
> shameful situation, I think.
>


I saw one more at an LBS the other day but I forgot the manufacturer
and model. I need to go back to that shop to find out.

I don't know if shameful is the right word. You basically have a
product (bicycle lights) where the market for a high end product at a
high price is very small, and where most consumers have no idea of the
difference between a $3 flasher and a $30 CatEye TL-DL1100.

If there was a well written standard requirement for bicycle lighting,
both front and rear, the situation would change. The problem is that
the existing standards are decades old, and are based not on what
would provide an adequate level of safety, but on what can be powered
from a dynamo (at least in parts of Europe).

As with many products, the responsibility is on the consumer to
understand that the government isn't going to protect them, and that
they need to buy the proper products. The problem is more that some
people believe that the lame minimum standards are actually
sufficient. You're pretty hard on Frank, but he's no different than
the typical non-scientist, non-engineer consumer that has no knowledge
of this type of product, and has no desire to spend time becoming
knowledgable. The thing is that it shouldn't be necessary to have to
become an expert on every product you buy!

> I really hoped that a hub dynamo would be so much better than sidewall
> dynamos but it is not so: the light output in my kinds of mild use is
> no higher. The truth is that a hub dynamo with even the best lamps
> designed for it doesn't make adequate light for any of my nighttime
> rides


They are often okay on known routes, or well lit streets, where you
won't encounter anything like you mentioned. There is only so much
power you can get out of a 3 watt dynamo, and efforts to convert the
native voltage and current into something more useful for LEDs results
in unacceptable losses in the conversion process.
 
On Mar 16, 5:02 pm, SMS <[email protected]> wrote:

> The highlight of my trip was meeting Joe Breeze as he rode in from his
> hotel on a Dahon, lent to him by Josh DaHon.


Oops, it's Josh Hon, not DaHon. I was talking to the wife of DaHon's
founder in the DaHon booth, about the whole name thing. She said they
were sure they would have won the lawsuit but at the time they didn't
want to spend money fighting the Hon office furniture company.
 
(PeteCresswell) wrote:
> What light bulbs/LED emitters are you powering with your
> Sturmey-Archer Dynohub?
>
> Front? Back?
>
> I did the thing with the bridge rectifier to convert it's output
> to DC - figuring I'd drive some LED emitters to get max light.
>
> But I'm having problems with the voltage regulator part. I get
> the circuit tuned to put out, say, 1.5 volts with 6 volts coming
> in from a 4-battery test source and then when I put another
> battery in series, the output voltage rises to 2.something.
>
> Probably something dumb I'm doing in the regulator, and I expect
> to do it right eventually.
>
> But it got me to wondering how voltage-tolerant light bulbs and
> LEDs are. My assumption going into this has been "not very",
> but I don't have much to base that on.
>
> Am I doing the voltage regulation piece in vain? Do I really
> need regulation? When I hang a voltmeter on the DC output it
> varies from 3 volts at a walking pace to about 20 volts on a
> downhill.
>
> So, bottom line, who is using what without toasting a lot of
> bulbs yet getting plenty light?


The combined characteristics of LEDs and dynamos make regulation difficult.

Dynamos generate a voltage proportional to speed by rotating a permanent
magnet in a coil. This eliminates the simplest method of regulation
(varying the field strength).

Incandescent bulbs increase resistance with power (filament
temperature). This makes them somewhat self-regulating. LEDs, on the
other hand, decrease resistance with temperature, making them somewhat
self-unregulating (prone to thermal runaway). LED characteristics are
even less favorable to regulation in that the resistance is extremely
non-linear, while filaments are approximately linear.

The two common ways to regulate are: putting a varying load in series
with the lamp load or switching the load (rapidly) in and out of the
circuit. The latter approach is significantly complicated by the
inductance of the dynamo, currents in inductors don't want to be
switched -- that's why dimmers on compact florescent lamps are not common.

These characteristics leave only the addition of a series dynamic load
as a simple regulator. Given the non-linearity of LED response and
thermal drifts, the regulator must sense current at relatively low
voltages. Unfortunately, even if done well, this approach results in
converting excess power into heat, rather than additional light. It's a
bit simplistic, but fairly accurate to think of a LED as a current
device and a dynamo as a voltage device and the problems that arise as
being from this fundamental mismatch.

An ideal way to match these two devices would be to have a lamp with
multiple LEDs (in series) and electronics that would sense the current
and keep it within limits by switching shunts across the individual
elements. Of course this complicates both the electronics and the optics.

The simple, "brute force" way would be to stack enough LEDs to match the
designed voltage range of the dynamo, spec'ing enough power handling in
the LEDs to handle peak dynamo output. This approach would take
advantage of the good efficiency and color of under-driven LEDs.

Practically speaking, if you used 3, 3W power LEDs (like the Cree),
you'd probably be able to handle the full output of the SA dynamo, but
the LEDs wouldn't start lighting until the dynamo was outputting 7.5V or
so. The inefficient approach would be to just use a single power LED and
pick an appropriate "ballast" resistor. You might be able to wire 2 or
more LEDs in parallel with separate ballast resistors in each "leg",
this would give you more light at low speeds, but you'd have to watch
out for current hogging and thermal runaway.

The ideal would be to simply have a single LED that was rated for the
full power output of the dynamo matched with a dynamo wound for the LED
characteristics. I'm not sure how close you can get to that with
available components.
 
On Mar 16, 9:55 pm, SMS <[email protected]> wrote:
> On Mar 16, 4:24 pm, Andre Jute <[email protected]> wrote:
>
> > That makes all of *three* generally available good tail lights... A
> > shameful situation, I think.

>
>
> I don't know if shameful is the right word. You basically have a
> product (bicycle lights) where the market for a high end product at a
> high price is very small, and where most consumers have no idea of the
> difference between a $3 flasher and a $30 CatEye TL-DL1100.
>
> If there was a well written standard requirement for bicycle lighting,
> both front and rear, the situation would change. The problem is that
> the existing standards are decades old, and are based not on what
> would provide an adequate level of safety, but on what can be powered
> from a dynamo (at least in parts of Europe).


Sorry, but it's pretty obvious that taillight standards, where they
exist, are not set by dynamo limitations. They are set by judgment
about adequate safety.

To explain: The typical bike generator is a 3 Watt unit. Typical
bulbs are chosen for 2.4 W headlights and 0.6 W taillights.

But first, if it were felt necessary to send more power to the rear,
that balance could have easily been changed. Authorities could have
called for 2W front, 1W rear. But they saw no need.

Second, 6 watt generators have been available (although not common)
for many years. If there were some rash of rear-end collisions in
countries where generators are dominant, the governments could have
simply mandated 6 W generators on all new bikes, with up to 3.6 W
devoted to taillights. But again, the road safety professionals in
those countries saw no such need.

Third, the invention (in the 1980s) of LED taillights was another
opportunity to increase brightness greatly, with zero technical
difficulty. VistaLight engineers, for example, could have simply
added more LEDs and reduced their battery life from, oh, 200 hours to
a still superb 150 hours. Obviously, they saw no need.

And the assertion by SMS and Jute that most taillights are inadequate
is typical of their unproven, yet dogmatic assertions. Scharf and I
have discussed this for years, and while he's stated that cyclists
should be using high-powered disaster-warning strobe lights, he has
not yet produced any evidence that legally lit cyclists are being mown
down at night. (Yes, the per-mile risk of fatality at night is
apparently higher, but indications are that the causes are a mix of
drunk cycling and zero lights or reflectors, rather than lights and
reflectors that are present but inadequate.)

In fact, all but two (IIRC) US states, and many other countries, have
no government requirements for taillights. That is, they judge
reflectors alone to be adequate in the rear. I do advocate taillights
for bikes, but I think it's safe to say that all modern taillights are
much more functional than mere reflectors - that is, they achieve the
level of conspicuity that most safety professionals have deemed
adequate.

> As with many products, the responsibility is on the consumer to
> understand that the government isn't going to protect them, and that
> they need to buy the proper products. The problem is more that some
> people believe that the lame minimum standards are actually
> sufficient.


A Scharfian proclamation of "lame" doesn't prove the standards are
insufficient!

In the past, I've cited papers testing the conspicuity of various
lights and reflectors for pedestrian and bicycle use. I won't dig
them out again now, since work is calling me, but IIRC, a minimal
reflector was visible to a motorist at something like a quarter mile,
and actively noticed by the motorist in _plenty_ of time for safe
passing.

And I think those of us who drive cars can verify this based on our
own experiences. How many times have you noticed simple LED blinky
taillights from 1/4 mile away? For me, it's countless times, in all
environments.

And how many times have you had to swerve your car at the last second
because you barely noticed you were approaching a legally lit cyclist
at night? For me, it's never. Scharf has mocked this in the past,
but really - if there's no evidence that legally lit cyclists are
being mown down or barely missed, then there's no evidence of a
problem.

The fact is, a bicyclist does NOT need extreme equipment to ride
safely at night. There will always be at least two groups who
disagree - the "cycling is dangerous!!!!" crew, and those who hope
(or, like Scharf, once hoped) to make money by selling expensive
lighting equipment.

But if these folks provide no evidence, they deserve no audience.

- Frank Krygowski
 
Andre Jute wrote:
> On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:
>> In article
>> <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com>,
>> Dan O <[email protected]> wrote:
>>
>>> On Mar 15, 2:09 pm, [email protected] wrote:
>>>> With one exception, I've never been in a place where a legally lit
>>>> cyclist is not sufficiently conspicuous. I've satisfied myself
>>>> literally hundreds of times that I'm even more visible at night
>>>> than I am in daytime.
>>> WTF?!?

>> WTF do you mean "WTF?!?"

>
> WTF do you mean WTF be means WTF?
>
>> A bicyclist at night with a headlight and
>> taillight is more conspicuous than a cyclist riding during the day. Way
>> more contrast between a bright light and a dark background.

>
> In total darkness, a condition that does not exist for most cycling
> situations. This is a juvenile form of argument, Tim.
>
> And even your basic argument is fatally undermined by the deficiency
> that the light must be pretty strong to be seen at the same maximum
> distance as a cyclist remains visiblein daylight


That would very much depend on what the cyclist was wearing and the
background he's viewed against. A BUMM standlight is easily visible
from more than a kilometer on a partially light cycle path at night, the
same is true in the day of course, but you have to find the cyclist first.

Sort of related, this is why I like flashing rear lights, because on the
road, at night, your rear light is just another one in a sea of red
lights, and the other ones usually come in pairs.
 
Andre Jute wrote:
> On Mar 16, 9:00 pm, SMS <[email protected]> wrote:


<snip>

There's also a rather ugly front light from Blackburn that I might us
> as a front flasher. But first I want to look into making something
> like the BUMM Fly IQ flash, so it can run off the dynohub and be there
> in cast I want to use it in steady mode. I'm also wondering if I can
> make the Cateye 1100 work off the dynohub for the rear.


Are you suggesting I shouldn't be running my IQ Fly off of my dynohub or
am I misunderstanding something here?

<snip>
 
Mon, 17 Mar 2008 15:20:23 GMT, Peter Cole:


>The combined characteristics of LEDs and dynamos make regulation difficult.


No.

>Dynamos generate a voltage proportional to speed by rotating a permanent
>magnet in a coil. This eliminates the simplest method of regulation
>(varying the field strength).


The main electrical charcteristic of hub generators is the limited current
output. You can simply hook a LED (with a current rating higher than that
of the generator) to the rectified generator output and you are done. If
you want more light, connect more LEDs in series.

>Incandescent bulbs increase resistance with power (filament
>temperature). This makes them somewhat self-regulating.


The current the bulb(s) can take, has to be the same as the max. output of
the generator.

>LEDs, on the
>other hand, decrease resistance with temperature, making them somewhat
>self-unregulating (prone to thermal runaway). LED characteristics are
>even less favorable to regulation in that the resistance is extremely
>non-linear, while filaments are approximately linear.


All LEDs have a specific max. current rating. Thats enough.

The nice thing about LEDs is: You can drive them with lower current
without loosing much light output - even the efficiency is higher at lower
current (just the opposite of filament bulbs).

>These characteristics leave only the addition of a series dynamic load
>as a simple regulator. Given the non-linearity of LED response and
>thermal drifts, the regulator must sense current at relatively low
>voltages.


Unneccessary.

> Unfortunately, even if done well, this approach results in
>converting excess power into heat, rather than additional light. It's a
>bit simplistic, but fairly accurate to think of a LED as a current
>device and a dynamo as a voltage device and the problems that arise as
>being from this fundamental mismatch.


No - both have a max. current as their dominant characteristic. E.g. a
Cree XRE 7090 LED should be run at not much more than 700mA. Modern hub
generators (Shimano DH-3N71, SON28, SRAM i-light, Novatec EDH-1, etc.)
will deliver not more than 600mA. The old Sturmey-Archer Dynohub delivered
at most around 350mA.

>
>The simple, "brute force" way would be to stack enough LEDs to match the
>designed voltage range of the dynamo,


There is no "voltage range". No-load voltage increases nearly linear with
rpm for the better hub dynamos. At 5 km/h (mounted in a 700c wheel) you
can measure 5 Volt rms, at 50 km/h around 50 Volt rms. But short-circuit
current nearly stays teh same independent from speed.

>Practically speaking, if you used 3, 3W power LEDs (like the Cree),
>you'd probably be able to handle the full output of the SA dynamo, but
>the LEDs wouldn't start lighting until the dynamo was outputting 7.5V or
>so. The inefficient approach would be to just use a single power LED and
>pick an appropriate "ballast" resistor.


You don't need any resistor. Try it!

>The ideal would be to simply have a single LED that was rated for the
>full power output


Replace "power" with "current".


Andreas
 
On Mar 17, 6:26 pm, Tosspot <[email protected]> wrote:
> Andre Jute wrote:
> > On Mar 16, 4:11 pm, Tim McNamara <[email protected]> wrote:
> >> In article
> >> <14a59561-15a4-4312-b04a-c7612f9cf...@s37g2000prg.googlegroups.com>,
> >>  Dan O <[email protected]> wrote:

>
> >>> On Mar 15, 2:09 pm, [email protected] wrote:
> >>>> With one exception, I've never been in a place where a legally lit
> >>>> cyclist is not sufficiently conspicuous.  I've satisfied myself
> >>>> literally hundreds of times that I'm even more visible at night
> >>>> than I am in daytime.
> >>> WTF?!?
> >> WTF do you mean "WTF?!?"  

>
> > WTF do you mean WTF be means WTF?

>
> >> A bicyclist at night with a headlight and
> >> taillight is more conspicuous than a cyclist riding during the day.  Way
> >> more contrast between a bright light and a dark background.

>
> > In total darkness, a condition that does not exist for most cycling
> > situations. This is a juvenile form of argument, Tim.

>
> > And even your basic argument is fatally undermined by the deficiency
> > that the light must be pretty strong to be seen at the same maximum
> > distance as a cyclist remains visiblein daylight

>
> That would very much depend on what the cyclist was wearing and the
> background he's viewed against.  A BUMM standlight is easily visible
> from more than a kilometer on a partially light cycle path at night, the
> same is true in the day of course, but you have to find the cyclist first.


Our tests were not on cyclepaths but on a busy road with a visually
busy background. We tested a common dayglow vest and a black jacket.
The black jacket in daylight didn't lose by much because a bicyclist
is actually quite visible simply as a moving block. Most cyclist round
here are also good about bright clothing.

> Sort of related, this is why I like flashing rear lights, because on the
> road, at night, your rear light is just another one in a sea of red
> lights, and the other ones usually come in pairs.


The flashing light wins hands-down in all situations. I suspect that
in another couple of generations commonly available flashing lights
might become too strong, as Jay Beattie has pointed out, and then
we'll be able to make a test to discover the optimum (and no doubt
Krygo and Andreas Oehler will complain that we're blinding poor
innocent motorists). At present I like my Cateye 1100 as the most
reasonably priced rear lamp that offers a measure of daylight safety
as sell.

I might add that I view the growth of LED rear lights in cars with
dismay; until quite recently a LED rear light served to identify a
bike, and I haven't given up on the dream of educating motorists to
take *more* care around a cyclist than around a car.

Andre Jute